35-watts from 2 Tubes?

[jbefumo]

This is an evolving design in my endless quest to do more with less.  I've posted the interim steps at various places on this board, but at this point, thought it might be more reasonable to give it its own topic -- I'm sure the moderator(s) will chastise me appropriately if I'm mistaken ...
Began with the notion of a Champ/Deluxe 'hybrid' - basically, a Champ (Single-Ended) with the power of a Tweed Deluxe (~22W).  Started it in mid-2014, but had to take a year off to deal with an annoying case of cancer, but got back to it as soon as I was up and about again. I had actually ordered a Deluxe Board from Doug Hoffman (which still remains unused), when someone kindly pointed me to DaGeezer's Little Wing, which really appealed to me, so I changed direction and built the V2 version shown in the attachment.  I had already decided to use a 6SL7GT in the front end, and since I didn't need a phase inverter, I added the cathode follower ... just because I had the extra triode.  I also moved the tone stack forward into the traditional Fender location, so it really bears minimal resemblance to its inspiration.
Built it with components on hand, hence the use of a 4-ohm speaker to reduce the effective impedance of the output transformer's primary.
In its original form, I used a LR8N3G regulator to establish the requisite G2 voltage, but that always bothered me since it added a lot of complexity to what was intended to be an exercise in simplicity.  Then it occurred to me that since the Class-A design's current draw doesn't vary significantly, on average, there was really no reason not to just use a big-ass resistor, which is what I ultimately did.  It's built and sounds quite good (will get around to shooting a video soon, I hope -- got to clean up my shop first 8^o  )!
After ruminating on the design for a while, it struck me that I could probably simplify it further, as shown in the V3 schematic  attached. In fact, I might try using a KT-150 and pushing it to 35 Watts or so (same reason some people climb big mountains, I suppose). It just seems like it would be so totally cool to make that kind of power with just two bottles!
Any potential 'gotchas', comments, criticisms, observations are solicited and appreciated!
 

[shooter]

fwiw, my KT88 SE build runs about 34Wpdis, 17W clean, 23W distorted, audio, I used *BIG* iron for both OT PS, and everything is nice n cool, well, sept the 88, but even she just takes it in stride.

[John]

Offhand, the only thing I see (and only because they talk about it a lot on here) is your SE OT is going to weigh a lot, and possibly cost quite a bit more than an equivalent PP OT would. I'm assuming you're talking 35 watts output power, not just dissipation?

[jbefumo]

I originally started this with a KT88 in mind, but then went to the KT-120.  BTW, my pair of Gold Lion KT88s are going into a 80W version of the Tweed OD Special -- stay tuned!

fwiw, my KT88 SE build runs about 34Wpdis, 17W clean, 23W distorted, audio, I used *BIG* iron for both OT PS, and everything is nice n cool, well, sept the 88, but even she just takes it in stride.

[jbefumo]

Ayup!  The 25W Edcor in the current version is pretty hefty.  I was thinking about a 50W Single-ended unit using a pair of 120s in parallel -- Talked to Mercury about it and they said they could wind me the OT, and it would be the size and weight of a 100W Marshall PP OT -- which, while heavy, isn't unreasonably so.  Might compensate somewhat by using a Neo speaker if I ever get around to building it, and yes, thinking Output power.  I guess the bottom line is I just can't get into spending the irreplaceable minutes of my life building something I could just go out and buy ...

Offhand, the only thing I see (and only because they talk about it a lot on here) is your SE OT is going to weigh a lot, and possibly cost quite a bit more than an equivalent PP OT would. I'm assuming you're talking 35 watts output power, not just dissipation?

[PRR]

Maximum G1 resistance for KT120 in self bias is 240K.

GRID 1 CIRCUIT RESISTANCE
FIXED BIAS - 51,000 OHMS
SELF BIAS - 240,000 OHMS
http://www.mif.pg.gda.pl/homepages/frank/sheets/084/k/KT120.pdf

Yes, you can probably cheat that to 517K, 9 times out of 10. But you are proposing to run the bottle HOT, so a "small" grid current could drift you into trouble.

12AX7 with 100K Rp can easily drive 200K (Fender design framework). Suggest 220K to ground and 22K in series with grid. (We'll overlook the 2K overage.)

[PRR]

I'd also think (I thot this before) that instead of a monster 500uFd first cap and taking OT/Plate there, you should go 100uFd+100r+100uFd. 2-stage filter. Gives more than double the midbass buzz reduction, which will be significant in an SE stage. As shown with raw first-cap at plate and choke-filtered B+ to screen etc, I suspect plate buzz will exceed all other sources.

33K to Screen sounds VERY high. The data value shows Ig2 as 14mA, which can not happen with 33K and 440V supply. 1K or so ought to be plenty here.

Your two PT primary wiring is drawn wrong but I assume you know what to do and are just tossing available symbols in.

_I_ would use resistor CT on heaters with a power tube this big. Shorts inside the HOT KT120 could stress the heater winding; couple 100r may shrug it off or smoke 24 cents of part.

Some 12AX7 will have less heater hum/buzz leakage if heater CT is returned to a small positive voltage. The KT120 cathode is appropriate and free.

[jbefumo]

THANK YOU!  Not sure where I got that big number -- probably made sense at some point in the cycle, and just remained.

Maximum G1 resistance for KT120 in self bias is 240K.

GRID 1 CIRCUIT RESISTANCE
FIXED BIAS - 51,000 OHMS
SELF BIAS - 240,000 OHMS
http://www.mif.pg.gda.pl/homepages/frank/sheets/084/k/KT120.pdf

Yes, you can probably cheat that to 517K, 9 times out of 10. But you are proposing to run the bottle HOT, so a "small" grid current could drift you into trouble.

12AX7 with 100K Rp can easily drive 200K (Fender design framework). Suggest 220K to ground and 22K in series with grid. (We'll overlook the 2K overage.)

[jbefumo]

Thanks! Yeah, I had that big cap on hand, so used it, but definitely appreciate this great advice.  I will tinker with this and see what happens.  I'm really stoked about this idea -- the main board would just be so dirt simple -- a dream for future maintenance.  XM52, in response to another post (http://el34world.com/Forum/index.php?topic=20350.msg214087#msg214087), pointed me toward a very cool 'tilt control circuit' (http://tubecad.com/2013/08/blog0267.htm), which got me thinking -- I might try building the amp with interchangeable 'tone' boards -- that tile control (no bypass), my usual Big Muff control built on a push-pull pot to bypass, and maybe even  a board using a concentric pot for tone and 'brick' solid state reverb ....  Would probably go with a Tweed Princeton cabinet (possibly extra deep to accommodate the big power tube, if necessary) and a 10" Alnico speaker ...  Could possibly even cram it into a Champ footprint ...

I'd also think (I thot this before) that instead of a monster 500uFd first cap and taking OT/Plate there, you should go 100uFd+100r+100uFd. 2-stage filter. Gives more than double the midbass buzz reduction, which will be significant in an SE stage. As shown with raw first-cap at plate and choke-filtered B+ to screen etc, I suspect plate buzz will exceed all other sources.

33K to Screen sounds VERY high. The data value shows Ig2 as 14mA, which can not happen with 33K and 440V supply. 1K or so ought to be plenty here.

Your two PT primary wiring is drawn wrong but I assume you know what to do and are just tossing available symbols in.

_I_ would use resistor CT on heaters with a power tube this big. Shorts inside the HOT KT120 could stress the heater winding; couple 100r may shrug it off or smoke 24 cents of part.

Some 12AX7 will have less heater hum/buzz leakage if heater CT is returned to a small positive voltage. The KT120 cathode is appropriate and free.

[jjasilli]

In its original form, I used a LR8N3G regulator to establish the requisite G2 voltage, but that always bothered me since it added a lot of complexity to what was intended to be an exercise in simplicity.  Then it occurred to me that since the Class-A design's current draw doesn't vary significantly, on average, there was really no reason not to just use a big-ass resistor, which is what I ultimately did.  It's built and sounds quite good




In Class A it's the plate current that remains nearly constant.  Tube specs (6550) show the screen current varies by the about the same factor of nearly 4X -- idle vs. full signal -- whether Class A or AB.  But the proof is in the pudding if it works and sounds good.

[jbefumo]

Oh -- still have lots to learn!  Well, next build I might look into having a transformer wound with a tap for that purpose?

In Class A it's the plate current that remains nearly constant.  Tube specs (6550) show the screen current varies by the about the same factor of nearly 4X whether Class A or AB.  But the proof is in the pudding if it works and sounds good.

[jjasilli]

That is an option, along with a UL tap as another alternative. 


My understanding is that the 6550 family of tubes (KT-88, now there's KT-90 & 120) were originally developed specifically for hi-fi use in UL operation.  My Stromberg-Carlson APH-1100 mono bloc PA amps, which approach hi-fi specs, use either a tap or separate winding (don't know which) to supply 600V to the plates & 300V to the screens of KT-88's.  (It's also possible for the screen supply winding to overwrap the higher voltage winding, in such a way as to cause ripple reduction to the screen supply by internal induction in the PT secondary.)  Anyway, wihtout separate supply volatges it's hard to get those 6550 screen - plate voltages to spec. 


But this is not hi-fi, so there's lot's of cheating in guitar amps.  In the final analysis, lots of 6550's are run successfully off regular PT's & sound good.

[jbefumo]

Actually, my OT has a UL tap -- don't quite recall why I didn't use it -- since I'll be opening it up to effect some of the above suggestions, I'll play around with that as well.  Thanks again!


Joe

That is an option, along with a UL tap as another alternative. 


My understanding is that the 6550 family of tubes (KT-88, now there's KT-90 & 120) were originally developed specifically for hi-fi use in UL operation.  My Stromberg-Carlson APH-1100 mono bloc PA amps, which approach hi-fi specs, use either a tap or separate winding (don't know which) to supply 600V to the plates & 300V to the screens of KT-88's.  (It's also possible for the screen supply winding to overwrap the higher voltage winding, in such a way as to cause ripple reduction to the screen supply by internal induction in the PT secondary.)  Anyway, wihtout separate supply volatges it's hard to get those 6550 screen - plate voltages to spec. 


But this is not hi-fi, so there's lot's of cheating in guitar amps.  In the final analysis, lots of 6550's are run successfully off regular PT's & sound good.

[PRR]

> My understanding is that the 6550 family ... developed specifically for hi-fi use in UL operation.

6550 is Ampex's VTR servo amp.

As the total lifetime market for Ampex VTRs was maybe a week's run at Tung-Sol, T-S promoted it for big audio amps.

UltraLinear was the hip fad of the day, so of course they had to give UL suggestions.

The tube itself does not care.

[shooter]

The tube itself does not care.

Ok, I'm duly confused, you mean the tube doesn't care if it's running in UL or more *traditional* pentode mode?
 my understanding, the UL tap, helps *control* the G2 swing, keeping it at a *fixed* range compared to the plate.  It helps keep the tube closer? to a *happy* point between full pentode and full triode operation.

[HotBluePlates]

The tube itself does not care.

Ok, I'm duly confused, you mean the tube doesn't care if it's running in UL or more *traditional* pentode mode?
 my understanding, the UL tap, helps *control* the G2 swing, keeping it at a *fixed* range compared to the plate.  It helps keep the tube closer? to a *happy* point between full pentode and full triode operation.

The tube itself does not care.

Triode mode is when the screen it attached to the plate, so as the plate voltage varies the screen voltage also varies in exactly the same way. Since a reduced screen voltage will also reduce plate current, the shape of the "plate curves" changes and the tube delivers reduced output power (particularly at plate current peaks when the plate voltage is swinging towards a minimum value). The screen is at the non-B+ end (the plate end) of the output transformer.

Pentode mode locks the screen to a fixed voltage. Now the plate current is largely unchanged with plate voltage swings, and the tube is capable of delivering more power. But it may also have more/different distortion than triode mode operation. The screen is at the B+ end of the output transformer (though often with an extra stage of filtering).

Ultralinear operation attaches the screen to a midpoint along the OT primary winding (in between the B+ end and the plate end). The screen voltage now varies, but not as much as in triode mode operation. Power output and distortion are a compromise value between triode and pentode operation, though screen power does contribute to total power output.

The tube works just fine in any of the modes; however, its performance is different in each mode.

[shooter]

The tube works just fine in any of the modes; however, its performance is different in each mode.

Thx, my understanding is still standing, just missed the nuance

[jbefumo]

Well, finally got it almost finished .... [I just ordered the two zener diodes shown in the attached schematic -- still using a 30k resistor for the G2].


Also, finally got around to recording some sound samples: https://youtu.be/i4_9cjQqKW0


Been playing bass through it quite a bit, and now I'm starting to think about building myself a bass amp....

Actually, my OT has a UL tap -- don't quite recall why I didn't use it -- since I'll be opening it up to effect some of the above suggestions, I'll play around with that as well.  Thanks again!


Joe

That is an option, along with a UL tap as another alternative. 


My understanding is that the 6550 family of tubes (KT-88, now there's KT-90 & 120) were originally developed specifically for hi-fi use in UL operation.  My Stromberg-Carlson APH-1100 mono bloc PA amps, which approach hi-fi specs, use either a tap or separate winding (don't know which) to supply 600V to the plates & 300V to the screens of KT-88's.  (It's also possible for the screen supply winding to overwrap the higher voltage winding, in such a way as to cause ripple reduction to the screen supply by internal induction in the PT secondary.)  Anyway, wihtout separate supply volatges it's hard to get those 6550 screen - plate voltages to spec. 


But this is not hi-fi, so there's lot's of cheating in guitar amps.  In the final analysis, lots of 6550's are run successfully off regular PT's & sound good.

[kagliostro]

I've listen the sample and I like it 

at the begin of the video there is a layout of the amp, can you post it here in the thread ?

Thanks

Franco

[jbefumo]

Thanks, and yeah I posted the schematic but not the layout because I wasn't sure it was up to date, but, with that proverbial grain of salt, here it is.


Funny thing: I was just thinking back to how, in younger days, it was either Marshall or it was crap.  I remember seeing The Dead at the Fillmore East (American Beauty), and thinking: "Don't they know how much better they'd sound if they had MARSHALLS?"  (Garcia was playing an ES-125 3/4, and it made such an impression that I have owned an ES140T 3/4, two 3/4 Melody Makers, and still hope to find the right 125 before I die, but I digress...)  Well, I'll digress again: Saw Pink Floyd on the first stop of their Meddle tour at Carnegie Hall -- they had stacks of Hiwatt heads and a wall of WEM bottoms and, of course, I was thinking HIWATTs????  Why not Marshalls?


Anyway, I guess I've come full circle.  I found myself increasingly scrutinizing the clean tones of amps first, and thinking about times when I might want power on tap, but without losing critical articulation.  I heard a quote a while ago but don't recall the source -- something to the effect of: "There's no such thing as a 'clean' pedal," and that really seemed to nail it. It's easy to make good distortion -- it's harder to make good clean.

I've listen the sample and I like it 

at the begin of the video there is a layout of the amp, can you post it here in the thread ?

Thanks

Franco

[kagliostro]

Thanks for sharing the layout

BTW, which kind of zener are you using ?

Franco

[jbefumo]

Glad you asked, because I'm just hand-waving at this point.  Here's what I have coming:


http://www.newark.com/webapp/wcs/stores/servlet/ProductDisplay?catalogId=15003&langId=-1&urlRequestType=Base&partNumber=31C4454&storeId=10194


http://www.newark.com/webapp/wcs/stores/servlet/ProductDisplay?catalogId=15003&langId=-1&urlRequestType=Base&partNumber=42K2994&storeId=10194


I just went for the highest power rating I could find in the voltages I needed, and figured I would just adjust the value of the resistor to make sure I feed them no more than they can dissipate, and rely on the zeners to even out any fluctuations ... near the mark?

Thanks for sharing the layout

BTW, which kind of zener are you using ?

Franco

[kagliostro]

Thanks again


Franco